On The Trail Of Autonomous Cars At CES 2017
ANALYSIS: Promising autonomous vehicle technology was everywhere you looked at CES 2017 in Las Vegas, but sorting out which promises might actually be fulfilled took some exploring.
Instead, the development was largely in the hands of independent suppliers, each of whom had some critical role to play in bringing safe, reliable, autonomous cars to market.
Savari, for example, has been working with Virginia Tech in those autonomous car tests on the Interstate highways around Washington, DC. In those tests, it’s beginning to look like autonomous cars are part of a solution to keep traffic moving down the highways during the DC rush hour rather than spending most of the time in a dead halt, which is what happens now.
Other companies are working on other aspects of the autonomous vehicle problem. Nvidia, for example, is using the raw computing power of its graphic processing units to handle the artificial intelligence requirements of autonomous driving, including the ability to apply deep learning to the task.
Major manufacturer support
AI and deep learning are critical to the safe operation of autonomous vehicles. Meanwhile, Nvidia and Mercedes-Benz announced a partnership between the companies with teams in Sunnyvale, California and Stuttgart, Germany, working to integrate artificial intelligence into cars.
Audi and Nvidia announced at CES 2017 keynote that they are working together on autonomous cars. That effort is similar to the deal between Nvidia and Mercedes collaboration, except that with Audi, the two companies announced plans to have autonomous cars on the road in 2020.
Of course there’s more to autonomous cars than the computers and sensors. I went over to see the folks at Robert Bosch, which is doing its own research on using artificial intelligence in control systems to safely guide the car down the road.
Bosch is working with a wide variety of car manufacturers, including Mercedes Benz and Audi, as well as a group of U.S. automakers, to integrate the self-driving capabilities into production vehicles.
Of course that’s just scratching the surface. What’s interesting about the move to autonomous vehicles is that the whole thing seems to be an industry-wide effort. Companies are working on standards so that communications between vehicles use a common language. They’re also testing to ensure that the AI systems react predictably to traffic and to ensure that vehicles interact reliably with highway systems in the same way so that traffic flow is predictable and safe.
I was surprised at the level of collaboration between companies, but it makes sense. The only way that cars can use an intelligent traffic network is if they all understand the same communications and the same language. That appears to be happening. And if it all comes together as they hope, it could mean the beginning of dramatically safer highways, significantly reduced costs in terms of money and eve time stuck in traffic.
Originally published on eWeek
